Identification of a cassette-borne
dfrA7
-like
gene that shows a 97 bp extension at the 3
ⴕ
-end
of the reading frame
J Antimicrob Chemother2001; 49:573–574 Kayode K. Ojoa,b, Diana E. Waturangia,c,
H. Akin Odelolab, Antonius Suwantoc,dand
Stefan Schwarza*
a
Institut für Tierzucht und Tierverhalten,
Bundesforschungsanstalt für Landwirtschaft (FAL), Dörnbergstrasse 25-27, 29223 Celle, Germany;
b
Department of Pharmaceutical Microbiology and Clinical Pharmacy, College of Medicine, University of Ibadan, Nigeria; cDepartment of Biology, Faculty of Science and Mathematics, IUC Biotechnology, Bogor Agricultural University, Bogor;
d
South East Asian Regional Center for Tropical Biology (SEAMEO-BIOTROP), Bogor, Indonesia
*Corresponding author. Tel: ⫹49-5141-384673; Fax: ⫹49-5141-381849;
E-mail: stefan.schwarz@fal.de
Sir,
Sulphonamides and trimethoprim are among the most frequently used antimicrobial drugs in Nigeria, where the combination of both drugs is used preferentially for the treatment of urinary tract infections. However, sulphon-amides and trimethoprim are commonly available, alone or in combination, over the counter in pharmacies in Nigeria without a doctor’s prescription. Self-medication and mis-use of these drugs favours the development of resistance to sulphonamides and trimethoprim in various bacterial pathogens.
During the course of a study on antimicrobial resistance in uropathogenicEscherichia colifrom humans in Nigeria, the multi-resistant E. colistrain UCH10386 was obtained from a 52-year-old female out-patient suffering from a urinary tract infection at the University College Hospital, Ibadan, Nigeria. This strain showed resistance to ampi-cillin, tetracycline, streptomycin, sulphonamides and tri-methoprim. All five resistance properties were associated with a plasmid of c.58 kb, designated pOJO10386, as con-firmed by transformation into the E. colilaboratory strain JM107 and conjugation into the rifampicin-resistant E. coli
HK225.
Extended PCR analyses were conducted using previously
described primers to determine the type of the resistance genes.1–3For this, plasmid DNA prepared from an E. coli
JM107:pOJO10386 transformant was used. These analyses revealed the presence of the tetracycline-resistance gene
tet(A), the two sulphonamide-resistance genes sul1 and
sul2, the streptomycin-resistance gene aadA2and a bla TEM-type ampicillin-resistance gene.
Furthermore, plasmid pOJO10386 was screened for the carriage of class I integrons by PCR using a Pwo poly-merase (Peqlab, Erlangen, Germany) with an extended proof-reading activity and primers that corresponded to the 5⬘-conserved segment (5⬘ -GGCATCCAAGCAGCA-AG-3⬘) and the 3⬘-conserved segment (5⬘ -AAGCAGAC-TTGACCTGA-3⬘) of class I integrons.3A single amplicon of c. 0.76 kb was obtained from plasmid pOJO10386, whereas the whole cell DNA of the recipient strain E. coli
JM107 did not yield an amplicon. This amplicon was cloned into pCR-BluntII-TOPO (Invitrogen, Groningen, The Netherlands) and the recombinant plasmid was trans-formed into E. coliTOP10. E. coliTOP10 cells that carried this recombinant plasmid showed resistance to trimetho-prim. Sequence analysis of the amplicon conducted on both strands confirmed the presence of a dfrA7-like gene that was part of a gene cassette (Figure).4
The amplicon comprised 764 bp; the 96 bp at the 5⬘-end corresponded exactly to part of the attI1 site of class I integrons, whereas the 56 bp at the 3⬘-end were identical to the qacEsequence of the 3⬘-conserved segment of class I integrons.5,6The 612 bp between these conserved segments represented a complete dfrA7 gene cassette that differed from known dfrA7cassettes in several aspects. Homology searches (http://www.ncbi.nlm.nih.gov/BLAST/; last accessed 7 November 2001) revealed three database entries of dfrA7 genes (accession nos AF139109, U31119 and X58425), all of which code for virtually identical proteins of 157 amino acids. In comparison with the other dfrA7genes, a single bp exchange (a씮c in dfrA7from pOJO10386) at
position 178 was found to cause a conservative amino acid exchange at codon 21 from isoleucine in all other DfrA7 variants to leucine in DfrA7 from pOJO10386. However, more important was a mutation at positions 558/559. Instead of the ‘aa’ that was present in all other dfrA7genes,
dfrA7 from pOJO10386 had a single ‘c’ at this position. Although the exchange of ‘a’ by ‘c’ at position 558 did not even change the amino acid (isoleucine) at codon 147, the loss of the second ‘a’ caused a frameshift mutation that extended the dfrA7 reading frame by 97 bp and, conse-quently, the length of the DfrA7 protein by 32 amino acids. This extension of the reading frame obviously did not
Correspondence
Journal of Antimicrobial Chemotherapy (2002) 49, 573–584
573 © 2002 The British Society for Antimicrobial Chemotherapy